<p>The diagnosis of unilateral congenital superior oblique muscle (SOM) palsy relies on a cluster of clinical findings due to the absence of a definitive confirmation test. This study aimed to compare qualitative and quantitative magnetic resonance imaging (MRI)-based assessments and to define a standardized quantitative threshold for diagnosing SOM atrophy. Patients with unilateral congenital SOM palsy who underwent orbital MRI, matched with age- and sex-paired controls were included. MRI images were assessed qualitatively (SOM atrophy) and quantitatively (surface and area measurements along the muscle axis) using multiplanar reconstructions. Ratios between the paralytic and non-paralytic SOM, as well as the inferior rectus muscle as a control, were analyzed. Inter-rater reliability, diagnostic thresholds, and diagnostic performance of quantitative measurements were evaluated. A total of 120 subjects (60 patients and 60 controls) were included. The inter-rater agreement for SOM atrophy detection was good to excellent for both qualitative and quantitative methods. Qualitative assessment identified SOM atrophy in 46.7–48.3% of cases, while quantitative analysis showed significantly smaller SOM surface and area on the affected side (p &lt; 0.001). The optimum diagnostic thresholds for SOM palsy were a surface ratio &lt; 0.9 or an area ratio &lt; 0.8 between the paralytic and non-paralytic SOM, correctly diagnosing 78.3% of cases while yielding a 10% false-positive rate. As a complement strategy to trochlear nerve visualization, a stepwise approach combining qualitative and quantitative assessments enhances the evaluation of SOM atrophy. The Eye-Surface-Area (ESA) method including three steps (1) Qualitative assessment (2) Surface measurement and (3) Area measurement provides a reliable method for identifying SOM atrophy.</p>